Ultrasensitive DNA detection based on target-triggered rolling circle amplification and fluorescent poly(thymine)-templated copper nanoparticles

We describe a novel strategy for the ultrasensitive detection of target DNA based on rolling circle amplification (RCA) coupled with fluorescent poly(thymine)-templated copper nanoparticles (poly T-CuNPs). In the presence of target DNA, a padlock DNA probe that consists of two regions: a target DNA-specific region and a poly(adenine) region, is circularized by the ligation reaction, and the subsequent RCA reaction is promoted to generate long, concatemeric, single-stranded DNA (ssDNA) with a lot of repetitive poly T sequences. As a result, a large number of poly T-CuNPs are formed, exhibiting a highly fluorescent signal. However, in the absence of target DNA or in the presence of non-specific target DNA, the padlock DNA probe is not circularized and the subsequent RCA is not executed, leading to no production of fluorescent poly T-CuNPs. With this simple strategy, we successfully analyzed the target DNA with the ultralow detection limit of 7.79 aM, a value that is 3 or 7 orders of magnitude lower than those of previous RCA-based fluorescent DNA detection strategies. In addition, the developed system was demonstrated to selectively discriminate non-specific target DNAs with one-base mismatch, suggesting potential application in the accurate diagnosis of single nucleotide polymorphisms or mutations.


Introduction
Nucleic acid detection is essential in the area of bioanalytical research including clinical diagnostics, biomedicine, and forensics since all organisms have their own unique genetic information, which could be a promising identication marker. [1][2][3][4] However, the amount of target nucleic acids in biological samples is too low to be reliably analyzed and thus amplication techniques are inevitably required. Until now, polymerase chain reaction (PCR) has been widely utilized to detect the lowabundance target nucleic acids, 5-8 but it has critical drawbacks including the requirement of thermocycling, technical expertise, and expensive instrumentation, which are not suitable for pointof-care-testing in facility-limited settings. [9][10][11] Recently, various isothermal nucleic acid amplication methods have been reported as promising alternatives to replace PCR-based assays. [12][13][14][15] Notably, rolling circle amplication (RCA) has received the special attention due to its unique feature to produce a long, concatemeric, single-stranded DNA (ssDNA) at the constant temperature. 16,17 In addition, various signaling methods for the effective analysis of RCA products have been suggested, which includes colorimetry, 18,19 electrochemistry, 20,21 and uorometry. 17,[22][23][24] Among these, the uorescent approaches have been thoroughly studied due to their reliability, simplicity, and sensitivity. The examples in this type of signaling strategies utilize uorescent dye-labeled dNTPs, 22 dye-labeled DNA probe, 24 molecular beacon, 17 uorescence resonance energy transfer (FRET) probe, 25 and intercalating dyes such as SYBR green. 23 Although each method provides the promising results for the effective analysis of RCA products, these strategies have several limitations. For example, the dyelabeled dNTPs signicantly reduce the amplication efficiency of RCA and the dye-labeled probes increase the overall assay cost. In addition, SYBR green dye that can be utilized in the realtime monitoring of RCA, inherently accompanies the high background signal due to its nonspecic intercalation property, which signicantly reduces the detection sensitivity. Therefore, a great incentive still exists for the development of sensitive and cost-effective methods for the target nucleic acids based on the uorescence monitoring of RCA reaction.
Towards this goal, we developed a simple, label-free, and ultrasensitive DNA detection strategy, which relies on targettriggered RCA and uorescent poly(thymine)-templated copper nanoparticles (poly T-CuNPs), a compelling alternative to conventional, organic uorescent dyes due to their outstanding spectroscopic and photophysical properties, low toxicity, and biocompatibility. 26 Notably, poly T-CuNPs are known to be more sensitive and quickly prepared as compared to DNA-templated silver nanoclusters (DNA-AgNCs), a wellknown uorescent signal reporter for the nucleic acid detection. [27][28][29][30][31] In principle, the presence of target DNA induces the circularization of a padlock DNA probe composed of a target DNA-specic region and a poly(adenine) region, and thus the subsequent RCA is promoted to generate a long, concatemeric ssDNA with a lot of repetitive poly T sequences. As a result, a large number of highly uorescent poly T-CuNPs are produced. Based on this simple strategy, the target DNA that encodes microRNA-141 (miR-141), a promising marker for human prostate cancer, was successfully analyzed with the ultra-high sensitivity and selectivity. 32,33 Experimental Materials All DNA sequences (Table S1 †) were synthesized from GenoTech (Daejeon, Korea) and puried by the standard desalting except 5 0 -end phosphorylated padlock DNA probe which was puried by polyacrylamide gel electrophoresis (PAGE). T4 DNA ligase, phi29 DNA polymerase, bovine serum albumin (BSA), and dNTPs were purchased from New England Biolabs (Beverly, MA, USA). Copper sulfate, sodium ascorbate, 3-(N-morpholino) propanesulfonic acid (MOPS), and sodium chloride were purchased from Sigma-Aldrich (St. Louis, MO, USA). All chemicals were of analytical grade and used without further purication.

Preparation of circularized DNA probe
The ligation solution (19 mL) composed of 1 mM of padlock DNA probe, 1Â T4 DNA ligase reaction buffer (50 mM Tris-HCl, 10 mM MgCl 2 , 1 mM ATP, 10 mM DTT, pH 7.5), and target DNA at varying concentrations or 10 pM of non-specic target DNAs was rst heated at 95 C for 5 min, and then cooled slowly to 16 C for 15 min to form a target DNA/padlock DNA probe complex. To this solution, 10 U of T4 DNA ligase (1 mL) was added, which was incubated for 60 min to make the circularized DNA probe. Finally, the ligation reaction was stopped by heating the solution at 65 C for 10 min and the target DNA/circularized DNA probe complex was stored at 4 C for further use.

Target DNA-triggered RCA reaction
The RCA reaction solution (40 mL) composed of as-prepared target DNA/circularized DNA probe complex, 500 mM of dNTPs, 4 mg of BSA, 1Â phi29 DNA polymerase buffer (50 mM Tris-HCl, 10 mM MgCl 2 , 10 mM (NH 4 ) 2 SO 4 , 4 mM DTT, pH 7.5), and 10 U of phi29 DNA polymerase was incubated at 30 C for 30 min and then terminated by heating the solution at 65 C for 10 min. The resulting product was stored at 4 C for further use.

Synthesis of poly T-CuNPs
Aer the target-triggered RCA reaction, the resulting product was mixed with 2 mM of sodium ascorbate and 0.5 mM of CuSO 4 in MOPS buffer (10 mM MOPS, 150 mM NaCl, pH 7.6), which was then incubated at room temperature in the dark for 5 min with gentle shaking to produce the poly T-CuNPs. The asprepared poly T-CuNPs were characterized by TEM analysis (Fig. S1 †) and their uorescence intensities were measured at the excitation wavelength of 340 nm.

Detection of target DNAs in human serum (1%)
The target DNA at varying concentrations and non-specic target DNAs with one-base mismatch (T1, T2, and T3) were spiked into the diluted human serum, which were subsequently analyzed using the same procedure to detect the target DNA in the buffer solution (target DNA-triggered RCA reaction and synthesis of poly T-CuNPs).

Instrumentation
The gel image was obtained by using Bio-Rad Gel Doc™ Ez Imager (Hercules, CA, USA) and the uorescence intensities were measured using a Tecan Innite M200 pro microplate reader (Mannedorf, Switzerland) and black, 384-well Greiner Bio-One microplates (ref: 781077). The TEM image of asprepared poly T-CuNPs was taken by using a eld-emission transmission electron microscopy (Tecnai, FEI, Netherlands) operating at an accelerating voltage of 300 kV.

Results and discussion
The overall procedure for the ultrasensitive DNA detection based on the target-triggered RCA coupled with the uorescent poly T-CuNPs is schematically depicted in Fig. 1. In this strategy, the padlock DNA probe which consists of two regions: one is complementary to target DNA and the other is 30-mer of poly(adenine) sequence, is utilized as a key detection component. In the presence of target DNA, the padlock DNA probe forms the target DNA/padlock DNA probe complex to be circularized by T4 DNA ligase. The RCA reaction is then promoted by phi29 DNA polymerase, leading to the generation of the long, concatemeric ssDNA. Importantly, the resulting RCA product contains a lot of repetitive poly T sequences, which consequently leads to the production of highly uorescent poly T-CuNPs. On the other hand, when target DNA is absent or non-specic target DNA is present, the padlock DNA probe is not circularized to perform the RCA reaction and thus uorescent poly T-CuNPs are not formed, which is manifested by the negligible uorescence signal. With this simple working principle, the target DNA is specically detected with the ultrahigh sensitivity.
We rst conrmed the detection feasibility of the proposed strategy by measuring the uorescence emission intensities from poly T-CuNPs. As envisioned in the design of the new strategy, the high uorescence signal from poly T-CuNPs with the emission maximum at 650 nm upon the excitation at 340 nm was observed in the presence of both padlock DNA probe and target DNA (1, Fig. 2(a)). In contrast, the weak uorescence signal was obtained when either of the two components (padlock DNA probe and target DNA) was not present (2 and 3, Fig. 2(a)). To further support the uorescence results in Fig. 2(a), the gel electrophoresis analysis of products obtained aer the target-triggered RCA reaction was executed. As shown in Fig. 2(b), the RCA products with high molecular weights were obtained near the loading well (>10.2 kbp) only in the presence of both padlock DNA probe and target DNA (lane 1, Fig. 2(b)), which was not observed in the negative controls where either the padlock DNA probe or target DNA was not included (lanes 2 and 3, Fig. 2(b)). Overall, the results obtained from both uorescence measurement and gel electrophoresis were well matched, which clearly veries the detection feasibility of the proposed method.
Next, we optimized the reaction conditions including the RCA reaction time and the concentration of Cu 2+ required for the ultrasensitive detection of target DNA. The RCA reaction time was rst optimized by measuring the uorescence signals of a single-stranded DNA specic uorescent dye, SYBR green II in a real-time signal acquisition mode. 34 The results in Fig. S2 † show that the uorescence signals remained quite low in the absence of target DNA, but it gradually increased to reach a plateau aer 30 min in the presence of target DNA. In addition, we optimized the concentration of Cu 2+ for the effective synthesis of uorescent CuNPs since dithiothreitol (DTT) essential for the RCA reaction can scavenge on Cu 2+ through its complexation with sulydryl groups in DTT. 35 As shown in Fig. S3, † the highest signal-to-noise ratio was obtained from 0.5 mM of Cu 2+ . Therefore, we selected 30 min of RCA reaction and 0.5 mM of Cu 2+ for further experiments.
Under the optimized reaction conditions, the detection sensitivity of our strategy was assessed by measuring the uorescence intensities at 650 nm at varying concentrations of target DNA. As shown in Fig. 3, the uorescence intensities increased with increasing concentrations of target DNA and an excellent linear relationship between DF 650 and logarithm of target DNA concentration was found in the range from 10 aM to 1 mM (R 2 ¼ 0.98) with the limit of detection of ca. 7.79 aM (3s/slope), a value that is signicantly lower or slightly higher than those from other RCA-based uorescent biosensors to detect the target DNA. 25,36,37 The ultra-high sensitivity of present   method is attributed to the synergistic effect of RCA and highly uorescent poly T-CuNPs. Importantly, it should be emphasized that this method is more sensitive than RCA-based one that relies on DNA-AgNCs, which implies that CuNPs-based uorescent signaling is more effective than AgNCs. 27 The detection specicity of our system was also evaluated by testing non-specic target DNAs with one base mismatch. The results in Fig. 4 show that only the target DNA led to the signicantly increased uorescence while the non-specic target DNAs with one base mismatch resulted in the negligible uorescence. These results clearly demonstrate that this system is highly selective, suggesting the potential application for the simple and sensitive SNP genotyping or mutation detection. [38][39][40][41][42] In addition, the practical applicability of the developed method was demonstrated by analyzing the mock clinical samples in which the target DNAs were spiked in human serum. [43][44][45][46] As shown in Fig. S4, † the target DNAs with varying concentrations were successfully analyzed in human serum with the high selectivity, conrming its potential for the detection of target DNAs in real, clinical samples.

Conclusions
In this study, we developed an ultrasensitive DNA detection strategy based on target-triggered RCA coupled with the uorescent poly T-CuNPs. By taking advantage of the high ampli-cation capability of RCA and highly emissive uorescence of poly T-CuNPs, we successfully detected the target DNA that encodes miR-141 with the high sensitivity and selectivity, which does not require any expensive labeling and complicated assay procedures. In addition, the system can be expanded to the detection of long target DNAs by employing an additional primer that initiates RCA (Fig. S5 †). Importantly, this method gave the limit of detection of 7.79 aM, a value that is 3 or 7 orders of magnitude lower than those of previous RCA-based uorescent DNA detection strategies. Moreover, the target DNA was successfully discriminated from non-specic target DNAs only with one-base mismatch, which suggests the future application of the proposed method for SNPs genotyping or mutation detection. We strongly believe that our strategy can be a powerful tool for the early diagnosis of specic nucleic acids related to infectious diseases or cancers.

Conflicts of interest
There are no conicts to declare. Fig. 4 Detection selectivity of the developed strategy. Fluorescence intensities at 650 nm (F 650 ) in the presence of target DNA and nonspecific target DNAs with one base mismatch (T1, T2, and T3) (Table S1 †). The final concentrations of padlock DNA probe, target DNA, and non-specific target DNAs are 1 mM, 10 pM, and 10 pM, respectively. All the experiments were performed in triplicate.